Claude Skills by HolobiomicsLab

Use when when you have a trained multitask model that accepts multiple input modalities (e.g., 1D NMR spectra in different nuclei or complementary analytical techniques) and you need to understand their relative importance for the downstream prediction task (e.g., molecular structure elucidation).

Use when a deep learning model for molecular structure prediction (e.g., NMR2Struct) has been trained and evaluated on a limited molecular size range (e.

Use when when you need to programmatically interface with a TensorFlow Serving model instance and must discover or validate the expected input names (e.

Use when when preparing to send peak data (1H and 13C NMR measurements) to a machine learning classification endpoint and you need to verify the current model's input/output names and schema, especially before implementing or updating code that constructs JSON payloads for the /api/smart3/search.

Use when before submitting peak data or other inputs to a machine learning classification API for the first time, after a model update, or if you encounter unexpected prediction errors. It is essential when the underlying model's input names or structure may change and require code updates.

Use when you have a pair of MS/MS spectra—one from a known compound and one from a structurally related modified (unknown) compound—and need to identify which atom(s) in the structure carry the modification.

Use when you have 1D ¹H and/or ¹³C NMR spectra (as preprocessed numerical arrays or peak lists) from an unknown organic molecule with ≤19 heavy atoms, and you need to recover its molecular formula and connectivity graph.

Use when you have 1D ¹H or ¹³C NMR spectra from an unknown organic compound and need to predict its molecular formula (e.g., C₆H₁₂O₂). The compound must contain ≤19 heavy (non-hydrogen) atoms. Use this as the first stage of a structure elucidation pipeline before predicting molecular connectivity.

Use when when you have encoded spectral features (from a CNN featurizer applied to 1D 1H and/or 13C NMR spectra) and a set of candidate molecular fragments predicted for a molecule with ≤19 heavy atoms, and you need to determine the correct connectivity and assembly order to recover the.

Use when after RDKit has generated multiple conformations for a molecule in an SDF or XYZ format, and you need to reduce the conformational ensemble to a tractable size (by energy-based ranking) before submitting to expensive quantum-chemical calculations (e.g., QUICK).

Use when when you have a training dataset of NMR spectra-derived molecular fragments encoded as token sequences, and you need to predict both the molecular formula and complete bond connectivity of unknown molecules.

Use when after generating candidate transformed structures from biotransformation rules and when you have MS/MS spectral feature data that you wish to organize into putative molecular families.

Use when you have 2D molecular structures (SMILES or SDF format) and need to create 3D conformer geometries as input to subsequent computational chemistry workflows, such as CCS prediction, molecular graph construction, or voxel-based property calculations.

Use when after an end-to-end neural model (CNN + transformer) has generated predicted molecular structures (formula and connectivity) from 1D NMR spectra.

Use when you have trained a multitask NMR-to-structure model and need to quantify its predictive accuracy on held-out test molecules.

Use when you have preprocessed 1D ¹H and/or ¹³C NMR spectra (as numerical arrays or feature tensors) from an unknown organic compound with ≤19 heavy atoms, and you need to predict both the molecular formula and the connectivity graph of the compound without prior structural hypotheses or reference.

Use when your goal is to assess whether a pretrained NMR2Struct model trained on molecules ≤19 heavy atoms can generalize to larger, more complex molecules, or whether accuracy degrades significantly.

Use when when you have executed the MultiModalSpectralTransformer architecture on a set of multi-modal spectroscopic inputs (NMR, HSQC, COSY, IR) and obtained predicted molecular structures, and you need to assess prediction accuracy and structural correctness against ground-truth or reference.

Use when after clustering pairwise mass differences into transformation groups from FT-ICR MS formula-assigned datasets, apply this skill when you have a catalog of transformation pairs (e.

Use when after cleaning and clustering LC-MS features in MS-CleanR, when you need to assign putative compound identities to a feature set with MS/MS spectra, and when candidate compounds may exist across multiple specialized databases (e.g., natural products, pharmaceuticals, contaminants).

Use when you have acquired complementary spectroscopic measurements (NMR, HSQC, COSY, IR) for the same molecular sample and need to combine them for structure elucidation.

Use when you have parsed spectral data from four distinct modalities (NMR, HSQC, COSY, IR) stored as separate numpy arrays or DataFrames with normalized chemical shifts and intensity ranges, and you need to create paired multimodal training records where each molecule's spectra across all.

Use when when you have downloaded raw spectroscopic datasets from multiple sources (NMR, HSQC, COSY, IR files) and need to combine them into a single coherent training corpus where each molecule is represented by all four modalities.

Use when you have computed raw p-values from partial Spearman correlations (or other univariate tests) between each metabolite in a SummarizedExperiment object and a phenotype of interest, adjusted for epidemiological confounders (e.

Use when you have a trained multitask model checkpoint and preprocessed spectral inputs (1D NMR spectra, 1H-only, 13C-only, or combined 1H+13C), and you need to generate simultaneous predictions of molecular formula and connectivity structure to quantify modality contributions, compare single vs..

Use when you have mwTab format files (Mass Spectrometry or Nuclear Magnetic Resonance experimental data from Metabolomics Workbench) that need to be loaded into Python for downstream analysis, validation, conversion to JSON, or programmatic manipulation of metadata and data sections.

Use when you have mwTab-formatted files from the Metabolomics Workbench containing MS or NMR experimental metadata and data blocks that need to be converted to JSON for integration with REST APIs, web applications, or downstream tools that expect JSON input.

Use when you have LC–MS/MS data processed through MZmine2 or MZmine3 and need to construct a feature quantification table for natural product discovery pipelines (e.g., INVENTA prioritization, GNPS networking, or metabolite annotation).

Use when you have LC-MS feature tables (m/z and retention time columns) and corresponding .mzXML or .mzML files, and you need to automatically classify whether extracted ion chromatograms represent genuine metabolomic features or false positives.

Use when you have newly assayed 1H-NMR metabolomics data from Nightingale Health (CSV or TSV format) and need to apply one or more published metabolic risk scores (Deelen et al. all-cause mortality, van den Akker MetaboAge, Würtz cardiovascular event risk, etc.).

Use when you have isolated one or more regions-of-interest (ROIs) from a proton NMR spectrum (defined as lower and upper chemical-shift bounds in ppm) and need to systematically generate a list of plausible metabolite assignments from a reference database.

Use when when you have raw INADEQUATE NMR spectral data (e.g., from a query sample or library) and need to extract peak coordinates and intensities as a first step toward identifying metabolite peak networks or comparing against a metabolite database.

Use when you have preprocessed 1H NMR spectral data with an unknown or ambiguous peak (e.g., at a specific chemical shift δ), and you need to determine its chemical identity.

Use when you have uploaded a pre-analytical data table containing sample metadata, processing delay timestamps (pre- and post-centrifugation), and NMR metabolomic measurements for a cohort of plasma or serum samples, and you need to assess how processing delays affect metabolite concentrations and.

Use when you have a trained multitask machine learning model for structure prediction, test set molecules with paired ¹H and ¹³C NMR spectra, and need to understand the marginal contribution of each NMR modality or justify multimodal input design.

Use when you have 2D NMR spectra (heteronuclear: HSQC, HMBC;

Use when when you have raw or processed 1D NMR spectra (FID or frequency-domain format) and need to extract individual peak identities and quantitative parameters in tabular form.

Use when you have proton (1H) and carbon-13 (13C) NMR peak measurements from a molecular sample and need to classify the molecule using the SMART 3 deep learning API. The peaks must be reformatted from their native instrument output into JSON before submission to the /api/smart3/search endpoint.

Use when you have extracted peak data (chemical shift values in ppm, multiplicities, integration) from a processed NMR spectrum (JCAMP, RAW, or mzML format) and need to match these peaks against known NMR signals to propose or confirm structural assignments.

Use when you have a CSV or table-formatted input spectral peak list (chemical shifts, intensities, metadata) destined for NMRformer metabolite identification, and you observe or suspect that the list contains noisy peaks or peaks with chemical shifts that do not match known reference standards.

Use when you have raw 1D NMR spectra (in NMRPipe or similar format) from biological samples (urine, worm, or other metabolomics experiments) and need to extract a structured peak table with chemical shifts and intensities for downstream metabolite identification or quantification.

Use when when you have preprocessed 1H NMR spectral data from flavor mixtures or similar compound identification tasks, and you need to identify which compounds are present.

Use when when you have raw or lightly processed 1D NMR spectra (¹H and/or ¹³C) from unknown organic compounds and need to extract latent spectral features prior to structure elucidation.

Use when when you have raw 1D NMR spectroscopic data (urine, worm, or other biological samples) in a format supported by NMRBox/NMRPipe and need to generate peak tables with automatic phasing, baseline correction, and spectral deconvolution without manual intervention.

Use when you have 2D NMR spectral data (heteronuclear and/or homonuclear experiments) from multiple samples in a natural products library or mixture, and you need to identify common structural scaffolds, correlate features across samples, or prioritize samples for further structural annotation.

Use when you have generated 1D FID time-domain data and Fourier-transformed frequency-domain 1H NMR spectra, or computed 2D COSY/HSQC correlation matrices, and need to write them to disk in a format that standard NMR software (e.

Use when when you have raw or preprocessed 1H NMR spectral tensors from flavor or chemical mixtures and need to generate high-level feature representations that capture both fine-grained local patterns (e.g., peak multiplet structure, coupling constants) and global spectral context (e.

Use when when working with raw 1H NMR FID data acquired on instruments like Bruker Avance spectrometers that require baseline correction, phase adjustment, and signal alignment before metabolite identification or statistical association testing can proceed reliably.

Use when you have raw NMR/IR/MS spectral data in vendor-specific (RAW), open (jcamp), or mass spectrometry (mzML) formats and need to parse, validate, and convert them to a standardized internal representation with extracted metadata and peak tables for visualization or further analysis in.

Use when you have raw 1D ¹H NMR spectroscopy output consisting of (1) a CSV file with chemical shift and intensity columns and (2) a TXT file listing detected peak chemical shifts, and you need to load and validate these into memory before passing them to a metabolite identification model like.